specification: API Commons Rate Limits
specificationVersion: '0.1'
schema: https://raw.githubusercontent.com/api-evangelist/interface-research/main/schema/api-commons.yml#/$defs/RateLimits
provider: Poolside
providerId: poolside-ai
created: '2026-07-02'
modified: '2026-07-02'
reconciled: false
tags:
- AI
- Foundation Models
- Enterprise
- On-Prem
- Rate Limiting
description: >-
  Poolside publishes no numeric public rate limits for its OpenAI-compatible
  API or its Identity Management / SCIM API. That is consistent with the
  product: each API only exists inside a customer's own Poolside deployment
  (Cloud VPC or on-prem/air-gapped), so throughput is bounded by the compute
  the customer provisions - either Poolside's pre-provisioned hardware (sized
  for "hundreds of developers" per node) or the customer's own NVIDIA GPU
  cluster - rather than by a shared multi-tenant quota Poolside enforces
  centrally.
notes: >-
  No per-key or per-minute request caps are documented as of the review date.
  Capacity planning is effectively a hardware-sizing conversation with
  Poolside at deployment time, not a published API rate-limit table.
sources:
- https://docs.poolside.ai/api/overview
- https://docs.poolside.ai/api/authentication
- https://poolside.ai/enterprise
limits:
- name: Chat / Completions Throughput
  scope: deployment
  metric: requests
  limit: hardware-bound (sized per deployment; not published)
  notes: Governed by the GPU capacity provisioned for the customer's deployment, not a Poolside-enforced quota.
- name: Identity Management / SCIM API Requests
  scope: account
  metric: requests
  limit: not published
  notes: No fixed numeric limit documented for /poolside/v1 or /scim endpoints.
- name: API Keys per Tenant
  scope: account
  metric: keys
  limit: not published
  notes: Developer keys are self-issued per user with deployment access; service-account keys require the tenant-admin role.
policies: []
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com